A screen for inhibitory peptides of hepatitis C virus identifies a novel entry inhibitor targeting E1 and E2

Hepatitis C virus (HCV) entry into hepatocytes is a multistep process that represents a promising target for antiviral intervention. The viral envelope protein E1E2 plays a critical role in HCV entry. In this study, we sought to identify peptide inhibitors of HCV by screening a library of overlapping peptides covering E1E2. Screening the peptide library identified several novel anti-HCV peptides. Four peptides from glycoprotein E2 were selected for further investigation. The 50% effective dose (ED50) was approximately 5 nM for each peptide. Our data indicated that these peptides inhibited HCV entry at the post-attachment step. Moreover, these peptides blocked cell-to-cell transmission of HCVcc and had broad-spectrum antiviral effects on HCVcc. These peptides exhibited combination inhibitory effects on HCVcc infection when combined with IFN-α2b or anti-CD81 antibody. Interestingly, we observed that E2-42 associated with E1 and E2. Our results indicate that E2-42 inhibits HCV entry via E1 and E2. These findings suggest a new avenue for HCV therapeutic development.

sequences that contain membrane-transiting motifs have been designed to inhibit virus entry into cells. For example, a peptide derived from the amino acids 710 to 725 of the HCV E2, inhibits HCV pseudoparticle infection 15 . The amphipathic α-helical peptide C5A derived from the membrane anchor domain of the HCV NS5A protein exhibits significant inhibitory effects against HCV infection in vitro 16 . Because the strategy of developing antiviral peptides based on viral membrane proteins is feasible, systematic screening of anti-HCV peptides from HCV E1E2 is promising.
In this study, we screened a peptide library from JFH1 E1E2 and identified four novel HCV inhibitory peptides functioning at the post-attachment and viral spread steps. We further determined that one peptide, named E2-42, interacted with E1 and E2. Overall, our work opens a new avenue for screening anti-HCV peptides and uncovers potential anti-HCV agents.

Results
Identification of antiviral peptides from HCV E1E2. To screen peptides that inhibit HCV infection, a peptide library of 126 overlapping peptides (15-mers offset by 10 amino acids for E1, by 11 amino acids for E2) covering the envelope protein E1E2 of the HCV JFH1 strain (GenBank no. AB047639) was designed. The peptides were incubated with Jc1P7NS2Gluc2a for 1 h, and then the virus was used to infect Huh7.5-CD81 cells.
Next, we investigated whether E2-42, E2-43, E2-78, and E2-79 inhibit HCV infection in a dose-dependent manner. As shown in Fig. 1C, the four peptides blocked HCV infection with a 50% inhibitory concentration (IC50) of 1-5 nM. Then, we evaluated the cytotoxicity of the four peptides. The four polypeptides had no effect on the viability of Huh7.5-CD81 cells at concentrations of 10 nM, indicating that the anti-HCV effects of E2-42, E2-43, E2-78, E2-79 were not due to toxicity. Effects of E1E2-derived peptides on HCV E1E2-mediated infection. HCVcc Jc1P7NS2Gluc2a was used to screen for antiviral activity. HCVcc was incubated with 10 nM peptides for 1 h prior to addition to Huh7.5/ CD81 cells, and Gaussia luciferase activity was assayed 72 h later. The results were normalized to DMSO-treated cells. (B) Characterization of the specificity of inhibitory peptides against HCVcc infection. HCVcc and VSV-Gpp (VSV-G-pseudotyped lentivirus) were incubated with 10 nM peptides for 1 h. Then, Huh7.5/CD81 cells were inoculated for 4 h with virus/peptide mixtures, and Gaussia luciferase activity was assayed 72 h later. IFN-α and anti-CD81 antibodies were used as positive controls, and DMSO was used as a negative control. (C) The ED50 values of peptides were calculated based on their inhibition of HCVcc activity at different peptide concentrations. (D) Cytotoxicity of peptides against Huh7.5/CD81 cells by MTT assay. Each bar represents the average of triplicate data points, with the standard deviation indicated by the error bars.  Fig. 2A). The virus/peptide mixtures were then used to infect Huh7.5-CD81 cells. The four peptides from E2 exerted a broad spectrum of antiviral effects against HCV but different inhibitory efficiencies for different genotypes of HCVcc. Peptides E2-42 and E2-43 had slightly higher efficiency than peptides E2-78 and E2-79. For genotypes 2a, 1a, and 1b, the inhibition rates were 80%, 75% to 80%, and 60% to 70%, respectively ( Fig. 2B,C,D and E). The genotype-dependent inhibitory effects may due to differences in amino acid composition among genotypes (Fig. 3A). Next, we synthesized conservative genotype 1a and 1b peptides and evaluated the inhibition ratio for the corresponding genotypes of HCVcc. As shown in Fig. 3B and C, the inhibitory effect of E2-79 from 1a or 1b was significantly decreased compared to that from genotype 2a, indicating that the differences in amino acid composition between genotypes in E2-79 may play an important role in its antiviral effect.

The four peptides inhibit HCV entry and cell-to-cell transmission.
To evaluate the effects of peptides on HCV replication, we used an HCV replicon cell line called JFH1 Luc-NS3-5B. After treatment with peptides for 72 h, the replicon cells were lysed to measure luciferase activity. As shown in Fig. 4A, the peptides did not inhibit HCV RNA replication. To analyze whether the peptides had any effect on HCV entry, the peptides were incubated with Huh7.5-CD81 cells at different time points of Jc1P7NS2Gluc infection. The results clearly indicated a decrease in HCVcc infection when the peptides were present during and after virus infection based on luciferase activity measurements (Fig. 4B) and HCV RNA levels (Fig. 4C). Furthermore, there was no effect if peptides were added to the cells as a pretreatment. These results suggest that the four peptides inhibit HCV entry at the post-attachment step.
HCV is also spread via cell-to-cell transmission. After infection with HCVcc, progeny viruses are transmitted from donor cells to acceptor cells, resulting in focal areas of infection transmission. To determine if E2-42, E2-43, E2-78, or E2-79 inhibit cell-to-cell transmission of HCV, Huh7.5-CD81 cells were infected with HCVcc and overlaid with media containing 50 nM peptides and 1% agarose. After 72 h of infection, the cells were stained with HCV core antibody. Focus sizes were measured by counting the number of cells per focus. As shown in Fig. 4D and E, cell-to-cell transmission was blocked by the four peptides. Taken together, these results demonstrate that the four peptides prevent HCV entry and cell-to-cell transmission.

Combination of inhibitory peptides with IFN-α2b or CD81 antibody.
To determine if the inhibitory peptides from E2 complement other anti-HCV drugs, we investigated their antiviral activity in combination with IFN-α2b or CD81 antibody. First, we established that the IC50 of IFN-α2b was 100 IU/mL and that the IC50 of CD81 antibody was 1 μg/mL (Fig. 5A). Then, peptides at 2.5 nM or 5 nM in combination with IFN-α2b or CD81 antibody were incubated with genotype 2a HCVcc (Jc1p7NS2Gluc2a) for 2 h and added to Huh7.5-CD81 cells. After 4 h of infection, the virus mixture was removed, and luciferase was measured 72 h later. The results illustrated that the four peptides exhibited combination effects with antiviral IFN-α2b and CD81 antibodies for the inhibition of HCVcc (Fig. 5B). Similar results were observed for genotype 1a HCVcc (H77/JFHGluc2a) and genotype 1b HCVcc (HB/JFHGluc2a) ( Fig. 5C and D). Taken together, these results indicate that the four peptides from E2 have potential for combination therapy with IFN-α2b and CD81 antibody.

E2-42 binds HCV envelope proteins but does not interact with HCV receptors.
On the surface of HCV particles, E1 forms a trimer in an E2-dependent manner, and both E1 and E2 are required for HCV entry 18 . Because E2-42, E2-43, E2-78 and E2-79 blocked HCV entry when the peptides were present during and after virus infection, we investigated whether these peptides interact with HCV envelope proteins using a GST pull-down assay. As shown in Fig. 6A and B, E1 and E2 associated with GST-E42 and E42-GST. Compared with E2-42 (residue 548-562), E2-43 (residue 552-566) is lack of sequence from 548 to 551, which is predicted to exposed to the surface and therefore a potential interaction site for E1 and E2 (Fig. 6C). However, the four  Table 1. Anti-HCV peptides in our screen.
peptides did not interact with HCV receptors (SRB1, claundin-1, CD81 and occludin) (Fig. 6D), consistent with the lack of antiviral activity of these four peptides when incubated with the cells before infection.

Discussion
This study describes a novel strategy to screen HCV inhibitors. We synthesized 126 peptides covering the HCV genotype 2a strain JFH-1 envelope glycoprotein E1E2. Each peptide consisted of 15 amino acids (AA) composed of two adjacent overlapping 10-AA (for E1) or 11-AA (for E2) regions. By screening the inhibition of HCVcc infection, we identified four novel anti-HCV peptides located in E2 (E2-42, E2-43, E2-78, E2-79). However, the screening might miss some inhibitory peptides because those peptides are not totally overlapped. These novel peptides strongly inhibited HCV independent of genotype. Interestingly, E2-42 associated with E1 and E2. Our results indicated that these four peptides inhibit HCV entry at the post-attachment step and block cell-to-cell transmission. Peptide E2-42 binding to E1 or E2 may also inhibit the late steps of HCV life cycle, such as HCV package and release. These inhibitory peptides might be used to treat HCV in combination with other anti-HCV agents. Viral entry is the first step of the HCV life cycle 14 . Our study demonstrated that the four peptides inhibit HCVcc entry at the post-attachment step in a pan-genotypic manner. E2-42, E2-43, E2-78, and E2-79 exhibited broad-spectrum inhibitory effects against different genotypes of HCVcc, including gene type 2a: Jc1p7NS2Gluc2a and J6CGluc2a; genotype 1a/2a: H77/JFHGluc2a; and genotype 1a/2a: HB/JFHGluc2a. The inhibition rate of these peptides for genotype 2a, genotype 1a/2a, genotype 1b/2a chimeric virus infection was 80%, 75% to 80%, and 60% to 70%, respectively. The inhibitory effects of peptides E2-42 and E2-43 were slightly greater than those of peptides E2-78 and E2-79.
In this study, we attempted to investigate the inhibitory mechanisms of E42/E42/E78/E79 on HCV. These peptides did not associate with HCV receptors, including CD81, SRB1, claudin 1, and occludin. Our data demonstrate that E42/E42/E78/E79 have no effects on HCV viral RNA replication. We speculate that E42 may block viral entry and spread via E1 and E2. Further studies are needed to understand the inhibitory mechanisms of E42/E42/ E78/E79.
The design of synthetic peptides for inhibition of viral infection is a classic strategy. For example, Enfuvirtide, a 36-AA peptide derived from the HIV virus envelope glycoprotein gp41 transmembrane region inhibits HIV-1 entry and has been used for clinical treatment of HIV-1-infected patients 19 . Several groups have identified HCV entry inhibitory peptides derived from E1E2 15,20 . Consistent with these findings, we observed that peptides from similar regions, such as E1-17, E1-18, E1-27, and E2-28, inhibit HCV entry, which suggests that our methods are feasible. During the preparation of our manuscript, a study reported that the region covering E2-78 and E2-79 plays a role in inhibiting HCV entry 21 . We narrowed down the inhibitory peptides in this area to 15 AA, shorter than the peptide in the previous study. Our screen identified E2-42 and E2-43 as two novel entry inhibitory peptides. We were able to identify E2-42 and E2-43 because we used 15-AA peptides, while the other group used longer peptides.
The four peptides exhibited significant combination inhibition of HCV infection in combination with IFN-α2b or anti-CD81 monoclonal antibody in a dose-dependent manner. In addition, the combination of inhibitory peptides and IFN-α2b or CD81 monoclonal antibody had a broad-spectrum inhibitory effect on HCV infection. In the future, we will optimize the structure and amino acid composition of the inhibitory peptides to improve their inhibition efficiency. We will also combine the peptides with DAAs to inhibit HCV infection and thus provide a reference for improving clinical HCV combination therapy strategies.
In conclusion, we identified and characterized four new E2-derived HCV inhibitors. These inhibitors blocked HCV entry and suppressed HCV spread in cell culture, suggesting that they are strong candidates for further antiviral drug development. Furthermore, we determined that one of these peptides (E2-42), interacts with E1/ E2; this interaction might play a role in the inhibition of HCV infection, opening a window to explore potential mechanisms of HCV entry.

Synthesis of peptides.
A total of 126 peptides derived from the HCV genotype 2a JFH1 strain (GenBank no. AB047639) protein E1E2 were synthesized, of which 37 peptides were from HCV E1 and 89 peptides were from E2. Each peptide contained 15 amino acids, with 10 (for E1) or 11 (E2) amino acids of overlap between adjacent peptides.
HCV infection assay. Huh7.5/CD81 cells were seeded in 48-well plates at a density of 1 × 10 4 /well before infection. The next day, the cells were infected with MOI = 0.05 focus forming unit (FFU)/cell in triplicate. Luciferase activity or HCV core expression level was assayed 72 h later.

TC G G C TG C AC G TG G A T G A A C T C C A C T G G T T T C A C C A A G A C T T G T G A G C T -3 ′ ( G S T -E 2 -C -4 3 f o r w a r d ) ; 5 ′ -C A C A A G T C T T G G T G A A A C C A G T G G A G T T C A T C C A C G T G C A G C C G A A C T C -
3′(GST-E2-C-43 reverse); 5′-GATGGAGGGTCTTCTCCACCTTCACCAGAACATCGTGGACGTAC A ATAC ATG G AG C T-3 ′ ( G ST-E 2 -C -7 8 for w ard ) ; 5 ′ -C C ATG TAT TG TAC G TC C AC G ATG T TC TG G TG A AG G TG G AG A AG AC C C TC -3 ′ ( G ST-E 2 -C -7 8 re ve r s e ) ; 5 ′ -G ATG G AG C T TC ACCAGAACATCGTGGACGTACAATACATGTATGGCCT CTCAGAGCT-3′(GST-E2-C-79 forward); 5′-CTGAGAGGCCATACATGTATTGTACG TCCACGATGTTCTGGTGAAGCTC-3′ (GST-E2-C-79 reverse). Constructs encoding CD81 were inserted into the EcoRI and XbaI sites of pCMV14-3XFLAG. The PCR primers were as follows: 5′-CCCAAGCTTATGTCCGGATCCTGGCTCCGCG-3′ (forward); 5′-GCTCTAGAGTAC ACGGAGCTGTTCCGG-3′ (forward). Constructs encoding occludin were inserted into the HindIII and XbaI sites of pCMV14-3XFLAG. The PCR primers were as follows: 5′-CCCAAGCTTATGTCATCCAGGCCTCTTG-3′ (forward) and 5′-GCTCTAGATGTTTTCTGTCTATCATAG-3′ (reverse).